Contributions of maternal emotional functioning to socialization of coping

Abstract

This study examined whether maternal emotional functioning—emotional awareness and depression—guides the coping suggestions mothers make to their children in the context of a common childhood stressor (peer victimization). Across two waves of a longitudinal study, 330 mothers and their second graders (mean age (M) = 7.95 years, SD = .33; 158 boys and 172 girls) completed questionnaires. Emotional awareness predicted more primary control engagement suggestions (directly addressing stress or emotions). Depression predicted fewer cognitive restructuring suggestions (thinking positively) and more cognitive avoidance suggestions (orienting thoughts away from stress). Interactive effects between maternal emotional functioning and child sex also emerged. This study elucidates the impact of mothers’ emotional functioning on how they teach their children to cope with stress.

Theory and research suggest that parents’ emotional functioning predicts individual differences in parenting (Adam, Gunnar, & Tanaka, 2004; Foster, Garber, & Durlak, 2008; for reviews, see Conley, Caldwell, Flynn, Dupre, & Rudolph, 2004; Goodman & Gotlib, 1999; Hammen, 2009). One key parenting task is to teach children how to cope with stress; indeed, research documents some important consequences of the coping suggestions parents make to their children (e.g., Abaied & Rudolph, 2010b, 2011; Kliewer, Fearnow, & Miller, 1996; Kliewer et al., 2006). Despite the importance of these socialization efforts, the determinants of individual differences in parents’ coping suggestions remain understudied (for an exception, see Abaied & Rudolph, 2010a). The goal of this study was to understand how maternal emotional functioning contributes to the coping suggestions mothers make to their children in times of stress, specifically when facing the common childhood stressor of peer victimization.

Conceptualization and consequences of socialization of coping

Socialization of coping refers to the messages parents communicate to their children about how to cope with stress (Abaied & Rudolph, 2010a, 2010b; Kliewer et al., 1996, 2006). The present research drew from Compas and colleagues’ theoretical framework (Connor-Smith, Compas, Wadsworth, Thomsen, & Saltzman, 2000; Compas, Connor-Smith, Saltzman, Thomsen, & Wadsworth, 2001) to differentiate between coping suggestions that encourage engagement versus disengagement. Engagement coping suggestions encourage children to orient themselves toward stressors or their cognitive or emotional reactions to stressors. Within this category, primary control engagement suggestions encourage children to change the stressor or manage their emotional reactions to stressors (e.g., problem solving, emotion regulation); secondary control engagement suggestions encourage children to adapt themselves to stressors (e.g., cognitive restructuring). In contrast, disengagement coping suggestions encourage children to orient themselves away from stressors either by staying away from the source of stress (i.e., behavioral avoidance) or by avoiding their cognitive or emotional reactions to stressors (i.e., cognitive avoidance).

Socialization of coping plays an important role in children’s development. Mothers’ coping suggestions are associated concurrently (Kliewer et al., 1996, 2006; Miller, Kliewer, Hepworth, & Sandler, 1994) and prospectively (Abaied & Rudolph, 2011) with the coping strategies children adopt in times of stress. Moreover, one study found a direct link between maternal socialization of coping and children’s subsequent emotional adjustment (Abaied & Rudolph, 2010b). Despite this documented importance of socialization of coping, little research has investigated how these parenting behaviors emerge. Some concurrent research links mothers’ own coping behaviors to the coping suggestions they make to their children (Kliewer et al., 1996, 2006; Miller et al., 1994). Mothers’ coping suggestions also are concurrently associated with family environment, social support, and socioeconomic status (Kliewer et al., 1996, 2006). Only one prospective study has examined the antecedents of socialization of coping, revealing that insecure maternal attachment predicted fewer engagement and more disengagement coping suggestions (Abaied & Rudolph, 2010a). Because attachment security is broadly related to emotion regulation (for a review, see Shaver & Mikulincer, 2007) and emotional intelligence (Kafetsios, 2004), this finding is consistent with the idea that mothers’ emotional functioning will guide their socialization of coping. However, additional research is needed to identify more proximal predictors of individual differences in coping suggestions.

This study aimed to advance knowledge of how maternal emotional functioning contributes to socialization of coping over time. We felt it was important to focus on socialization of coping within a particular context. Just as there might be contextual variation in the coping strategies individuals use (e.g., Folkman, 1984) and in the consequences of coping suggestions for children’s adjustment (Abaied & Rudolph, 2010b; Compas et al., 2001; Compas, Malcarne, & Fondacaro, 1988), mothers’ coping suggestions may depend on the type of stress encountered by children. We specifically examined mothers’ socialization of coping in the context of peer victimization during elementary school. This context was selected for several reasons. First, prior research indicates that mothers’ socialization of coping predicts children’s coping (Abaied & Rudolph, 2011) and adjustment (Abaied & Rudolph, 2010b) when children face interpersonal stress. Second, peer victimization is a salient and common interpersonal stressor (Hanish & Guerra, 2002; Smith & Shu, 2000; Solberg & Olweus, 2003) that can have lasting adverse effects when severe or persistent (Rudolph, Troop-Gordon, Hessel, & Schmidt, 2011). Finally, children’s self-regulatory abilities mature in middle childhood (Calkins & Keane, 2009), and increasing levels of peer victimization and emerging self-regulation may stimulate mothers to begin teaching children how to cope independently rather than merely soothing children in times of stress. We anticipated that this developmental change would provide an ideal context in which to examine shifts in mothers’ coping suggestions over time. We specifically focused on two dimensions of maternal emotional functioning—emotional awareness and depression—that we believed would guide mothers’ coping suggestions.

Emotional awareness as an antecedent of socialization of coping

Emotional awareness refers to how individuals perceive and process emotions. Theory and research have conceptualized emotional awareness—and similar constructs such as emotional intelligence and emotional competence—using varying combinations of traits and skills. Across these conceptualizations, four commonly investigated aspects of emotional awareness include attention (inclination to reflect on emotions), clarity (ability to identify, understand, and distinguish one’s emotions), description (ability to put emotions into words), and expression (inclination to express emotions outwardly; Ciarrochi, Scott, Deane, & Heaven, 2003; Gohm & Clore, 2002; Salovey, Mayer, Goldman, Turvey, & Palfai, 1995; Salovey, Woolery, Stroud, & Epel, 2002).

Emotional awareness can influence how individuals cope with stress and respond to others’ distress. Individuals with high emotional clarity require fewer resources to process initial emotional responses to stress, allowing them to engage in resource-intensive coping strategies, such as actively planning ways to address the source of stress, reframing experiences in a positive way, and focusing on growth (Gohm & Clore, 2002). Individuals who attend to and express emotions also engage in more active coping responses, such as seeking social support (Gohm & Clore, 2002). Moreover, emotional awareness is associated with lower physiological arousal in response to repeated stress, including habituated cortisol release and lower systolic blood pressure (Salovey et al., 2002). Lower levels of physiological arousal following stress may free resources and allow individuals to engage in active coping. Finally, emotional awareness is associated with higher levels of empathy, including empathic and emotional responses to others’ distress (Salovey et al., 2002). Thus, emotionally aware individuals are likely to respond to stress using active engagement rather than disengagement, to view engagement as a beneficial coping strategy, and to be responsive to others’ distress.

Because emotional awareness can influence the coping strategies individuals use in response to their own stressors as well as their responses to others’ distress, we hypothesized that emotional awareness would predict how mothers socialize their children to cope. Because mothers high in emotional awareness attend to and understand emotion, are emotionally expressive, and themselves use more active forms of coping (Gohm & Clore, 2002), they may be more motivated and able to suggest coping strategies that encourage their children to address the source of stress or their emotional reactions (i.e., primary control engagement) or to adapt to stressors by focusing on growth or thinking positively (i.e., cognitive restructuring). Moreover, mothers high in emotional awareness may view cognitive avoidance (orienting thoughts away from the stress) or behavioral avoidance (staying away from the source of stress) as less effective strategies because they involve ignoring emotional responses to stress; consequently, they may be unlikely to suggest such strategies to their children. Thus, we expected maternal emotional awareness to predict more primary control engagement and cognitive restructuring suggestions and fewer cognitive and behavioral avoidance suggestions.

Maternal depression as an antecedent of socialization of coping

While emotional awareness may enhance mothers’ resources, making them more motivated and able to encourage active, resource-intensive coping strategies, maternal depression may deplete mothers’ resources, limiting their inclination and ability to suggest such strategies. Depressed and nondepressed mothers differ in their coping, with depressed mothers using fewer active strategies such as help and support seeking compared to nondepressed mothers (Churchill, Villareale, Monaghan, Sharp, & Kiekhefer, 2010). Because mothers’ own coping behavior is related to the strategies that they suggest to their children (Kliewer et al., 1996, 2006; Miller et al., 1994), maternal depression may predict fewer active coping suggestions directed toward children.

The behavioral, motivational, and cognitive sequelae of depression also may guide mothers’ socialization of coping. Depressed mothers tend to disengage from their children (for reviews, see Hammen, 2009; Lovejoy, Graczyk, O’Hare, & Neuman, 2000). Coping suggestions that encourage problem solving, regulating emotions, or reappraising experiences require mothers to provide more assistance and to become more engaged in the child’s stress and emotions than suggestions that encourage avoiding stressors and cognitive or emotional responses. Lack of energy and initiative also may make avoidance suggestions more appealing to depressed mothers. A sense of helplessness may make depressed mothers less likely to encourage problem solving and emotion regulation, and a sense of pessimism may make them less likely to help children reframe stress in a positive light. Indeed, when working with children on goal-oriented tasks, depressed mothers show less warmth, involvement, and constructive guidance (Foster, Garber, & Durlak, 2008) and are less comfortable with the teaching role (Goldsmith & Rogoff, 1995) than nondepressed mothers. Overall, we expected maternal depression to predict fewer primary control engagement and cognitive restructuring suggestions and more cognitive and behavioral avoidance suggestions.

Alternative predictors of socialization of coping

We examined the extent to which maternal emotional functioning predicted socialization of coping after adjusting for several alternative viable predictors. Because prior research links demographic characteristics and aspects of the family environment with mothers’ coping suggestions (Kliewer et al., 1996), in this study, we considered mothers’ marital status and family income. These factors have been found to influence mothers’ stress level and their ability to engage in warm, involved parenting practices (e.g., Conger et al., 1992; Simons, Lorenz, Conger, & Wu, 1992; Webster-Stratton, 1990). Mothers who are married or who have a higher family income may be able to devote time and effort toward helping their children use resource-intensive coping strategies such as primary control engagement and may be less likely to suggest cognitive or behavioral avoidance.

Drawing from theory and research suggesting that children can elicit parenting behaviors (Bell, 1968; Belsky, 1984; for a review, see Karraker & Coleman, 2005), we also examined the predictive contributions of children’s temperament and peer victimization experiences. Mothers of children high in temperamental negative emotionality may find that their child has difficulty addressing stressors directly, managing their emotional responses, or reframing stressors. Indeed, in a sample of children coping with cancer, child negative emotionality was associated with lower use of such coping strategies as reported by mothers (Miller et al., 2009). Thus, mothers of children high in negative emotionality may be more likely to encourage their child to use cognitive and behavioral avoidance. Mothers also may recommend different coping strategies depending on children’s level of exposure to peer victimization. Although this has not been examined directly, parents of victims and nonvictims do exhibit different parenting behaviors; for example, parents of victims are more likely to be overprotective or coercive than parents of nonvictims (Finnegan, Hodges, & Perry, 1998; Rigby, Slee, & Cunningham, 1999). Regarding coping suggestions, parents of children who experience frequent victimization might prefer to suggest avoidance rather than engagement with stress in attempts to minimize their child’s exposure to verbal or physical harm from peers. Thus, we examined the unique contribution of maternal emotional functioning to socialization of coping after considering family demographics and child characteristics.

We also considered the possibility of sex differences in mothers’ socialization of coping (main effects) and in the contribution of maternal emotional functioning to socialization of coping (interactive effects). Differences in the coping inclinations, needs, or responses of girls and boys and in mothers’ conceptions of gender-appropriate behavior may lead mothers to make different coping suggestions to sons and daughters. Specifically, girls are more inclined to use problem solving, support seeking, and emotion regulation (i.e., primary control) coping strategies (Donaldson, Prinstein, Donofsky, & Spirito, 2000), whereas boys tend to use more avoidance coping strategies (Eschenbeck, Kohlmann, & Lohaus, 2007; Compas et al., 1988; for an exception, see DeBoo & Spiering, 2010) and are more likely to respond with aggression when parents encourage them to use engagement coping (Abaied & Rudolph, 2010a). Conceptions of gender-appropriate behavior also may lead mothers to encourage emotional expression and discussion in girls (Cassano, Zeman, & Perry-Parrish, 2007; Fivush, Brotman, Buckner, & Goodman, 2000). Thus, mothers may be more likely to encourage primary control engagement in girls than in boys and to encourage avoidance in boys than in girls (Miller et al., 1994).

Additionally, we anticipated that maternal emotional functioning may heighten or impair mothers’ ability to provide coping suggestions that are contingent on children’s gender. Compared to nondepressed mothers, depressed mothers are less able to gauge their children’s thoughts and feelings (Coyne, Low, Miller, Seifer, & Dickstein, 2007) and are less likely to adapt their behavior to children’s cues and abilities (e.g., Hoffman & Drotar, 1991; Johnston, Murray, Hinshaw, Pelham, & Hoza, 2002). Similar to nondepressed mothers, mothers with high levels of emotional awareness may be better able to monitor their children’s abilities and needs and adapt their behavior accordingly. Thus, we anticipated that mothers with sufficient resources (high emotional awareness or low depression) would be more likely than mothers with insufficient resources (low emotional awareness or high depression) to show contingent parenting in which they modify their coping suggestions based on beliefs about what strategies are more typical or effective in girls versus boys (e.g., engagement in girls and disengagement in boys).

Method

Participants and procedures

Participants included 330 mothers and their second graders (mean age (M) = 7.95 years; SD = .33; 158 boys, 172 girls; 75% White, 14% African-American, 7% Asian, and 5% others). Families represented a range of family annual income levels (22% earning <$30,000; 42% earning between $30,000 and $75,000; and 35% earning >$75,000) and education levels (of the maternal caregivers, 19% completed some graduate school, a master’s, or professional degree; 23% earned a bachelor’s degree; 44% had completed some college or an associate’s degree; and 14% earned a high school degree or less.).

Participants were drawn from a sample of families participating in a longitudinal study based on the availability of parent data. Participating families were recruited from several small urban and rural school districts in the Midwest. Schools were selected to represent the ethnic and socioeconomic diversity of the region. However, as noted below, the participating mothers were somewhat less diverse than the participating sample of children. Consent forms were sent home through schools and distributed at parent–teacher conferences. Parents provided written consent for participation; children provided oral assent. Eighty percent (n = 576) of eligible children received consent and participated in Wave 1 (W₁) of the study. Participants and nonparticipants at W₁ did not differ in age (t(723) = .63, NS), sex (χ²(1) = .15, NS), ethnicity (χ²(1) = .59, NS), or school lunch status (full payment vs. subsidized; χ²(1) = .35, NS). W₁ and Wave 2 (W₂) data were available for 330 maternal caregivers (325 biological mothers, 3 grandmothers, 1 adoptive mother, and 1 step-mother). Data provided by paternal caregivers (n = 10) or by different caregivers at the two waves (n = 14) were not included. Children whose mothers participated at both waves versus neither or one wave did not differ in age (t(564) = 1.42, NS) or sex (χ²(1) = .32, NS), but children of mothers who did not participate at both waves were more likely to be minorities (χ²(1) = 21.28, p < .001) and recipients of subsidized school lunches (χ²(1) = 12.92, p < .001). Mothers who participated in one versus both waves did not differ in socialization of coping (ts(423) < 1.44, NS) or emotional awareness (t(423) = −1.05, NS), but mothers participating at both waves had lower levels of W₁ depression than mothers who did not (t(423) = 2.88, p < .01).

Mothers and children completed questionnaires when children were in second grade, and mothers completed questionnaires again when children were in fourth grade. Mothers received and returned questionnaires by mail or home visits. Children completed questionnaires in small groups at school. Mothers received monetary compensation and children received a small gift.

Measures

Table 1 presents descriptive statistics. The measures had adequate reliability, with the exception of a moderate reliability coefficient for the W₁ cognitive avoidance subscale.

Table 1.

Descriptive statistics (N = 330).

Measure	M	SD	α
W1 maternal emotional awareness	3.89	.54	.88
W1 maternal depression	2.19	.60	.93
W1 child negative emotionality	2.59	.56	.91
W1 child victimization	2.13	.78	.91
W1 primary control engagement	.29	.05	.69
W1 cognitive restructuring	.38	.05	.74
W1 cognitive avoidance	.14	.04	.57
W1 behavioral avoidance	.12	.03	.74
W2 primary control engagement	.32	.05	.76
W2 cognitive restructuring	.21	.04	.74
W2 cognitive avoidance	.29	.05	.87
W2 behavioral avoidance	.17	.04	.81

W₁: Wave 1; W₂: Wave 2.

Maternal socialization of coping

At W₁ and W₂, mothers completed a measure that taps suggestions mothers make to their children about how to cope with peer victimization. Mothers were provided the prompt “When other kids are mean to my child, I encourage my child to … ” and rated how much they use various coping suggestions on a 5-point scale (Not at all to Very much). This measure assesses mothers’ coping suggestions in accordance with the engagement–disengagement framework of voluntary responses to stress (Compas et al., 2001). Items were derived from Abaied and Rudolph’s (2010b) Socialization of Coping Questionnaire and from the Responses to Stress Questionnaire (RSQ; Connor-Smith et al., 2000). Some items from Abaied and Rudolph’s (2010b) questionnaire were shortened or reworded for clarity; RSQ items were reworded to capture coping suggestions rather than responses. Items were added to assess a more diverse array of coping suggestions than the original measure, allowing for more fine-grained distinctions among types of suggestions. The W₁ measure included 17 items. For W₂, two items were modified slightly for clarity, two items were dropped due to low factor loadings, and eight items were added across subscales to improve reliability; this yielded a 23-item measure.

Items captured four conceptual domains of coping suggestions: Primary Control Engagement (six items; e.g., “Do something to try to fix the problem or take action to change things.”; “Discuss his/her feelings with me or others.”), Cognitive Restructuring (five items; e.g., “Look for something good in what is happening.”), Cognitive Avoidance (eight items; e.g., “Not focus on the problem.”), and Behavioral Avoidance (four items; e.g., “Keep away from things related to the problem.”). A confirmatory factor analysis was conducted to examine whether this four-factor model adequately represented the structure of socialization of coping suggestions. Results indicated that the model provided an excellent fit, χ²(191) = 346.33, NS, χ²/df = 1.81, comparative fit index (CFI) = .95, incremental fit index (IFI) = .95, and root mean square error of approximation (RMSEA) = .05. Standardized regression weights ranged from .28 to .88 (average = .62). Proportion scores were computed by dividing the total score of the items comprising each factor by the total score of the measure. Proportion scores correct for base-rate differences in endorsement of responses to stress (Compas et al., 2001; Connor-Smith et al., 2000). Supporting the validity of parent reports of socialization of coping, these reports are significantly associated with children’s reports of their own responses to stress (Abaied & Rudolph, 2011; Kliewer et al., 1996; Shipman & Zeman, 2001) and teacher reports of children’s coping (Eisenberg, Fabes, & Murphy, 1996).

Maternal emotional awareness

At W₁, mothers completed the Emotional Awareness Questionnaire. This measure draws from several established measures of emotional awareness, including the Toronto Alexithymia Scale (Bagby, Parker, & Taylor, 1994), the Trait Meta-Mood Scale (Salovey et al., 1995), and the Emotional Expressiveness Scale (Kring, Smith, & Neale, 1994), to capture diverse aspects of emotional experience proposed to influence coping (for a review, see Gohm & Clore, 2002). The 20-item measure taps four aspects of emotional awareness: attention to one’s emotions (five items; e.g., “I often think about my feelings.”), clarity of one’s emotions (five items; e.g., “I almost always know exactly how I feel.”), ability to describe one’s emotions (five items; e.g., “I think it is easy to describe my feelings.”), and inclination toward expressing one’s emotions (five items; e.g., “I show my feelings to other people.”). Mothers rated how much they agreed with each item on a 5-point scale (Not at all to Very much). Although the measure encompasses several aspects of emotional awareness, these dimensions were conceptually related and sizably correlated (rs = .29–.68, ps < .001; average r = .48, after dropping one attention item with low intercorrelations). A confirmatory factor analysis indicated that a one-factor model provided an excellent fit to the data (χ²(113) = 150.06, NS, χ²/df = 1.33, CFI = .98, IFI = .98, and RMSEA = .03). Standardized regression weights ranged from .37 to .70 (average = .51). Thus, scores were computed as the means of the items, with higher scores indicating more emotional awareness. Adult reports of emotional awareness have strong reliability (i.e., internal consistency and stability) and demonstrate convergent and discriminant validity (Bagby et al., 1994; Kring et al., 1994; Salovey et al., 1995).

Maternal anhedonic depression

At W₁, mothers completed the anhedonic depression subscale of the Mood and Anxiety Symptom Questionnaire (Watson et al., 1995). This 22-item subscale assesses (lack of) positive affect and loss of interest (e.g., “I felt like nothing was very enjoyable.”). Mothers rated how much they experienced each symptom on a 5-point scale (Not at all to Extremely). Scores were computed as the mean of the items. This measure was selected to provide a pure measure of depression rather than a measure of more general emotional distress. Convergent and discriminant validity have been established, and evidence indicates that the anhedonic depression subscale specifically assesses depression rather than general distress or anxiety (Watson et al., 1995). The mean anhedonic depression score (Table 1) corresponds to a response of “A Little Bit” on the rating scale, as would be expected in a nonselected community sample and similar to the reported score in a comparable nonclinical female sample in which the measure was developed (M = 2.36, Watson et al., 1995).

Family demographics

At W₁, mothers reported on their marital status (married; single, never married; single, divorced; separated; other) and family income (1 = <14.999 to 7 = >90,000). For ease of interpretation, marital status was coded as 0 = single, separated, other; 1 = married in the analyses.

Child negative emotionality

At W₁, mothers completed three subscales of the Temperament in Middle Childhood Questionnaire (Simonds, Kieras, Rueda, & Rothbart, 2007) that tapped negative emotionality: soothability (8 items, reverse scored; e.g., “My child cheers up quickly”), anger (6 items; e.g., “My child gets angry when he/she makes a mistake”), and sadness (10 items; e.g., “My child’s feelings are hurt easily”). Parents rated how true each item was of their child on a 5-point scale (Almost always untrue to Almost always true). The three subscales were strongly intercorrelated (rs ≥ .60, ps < .001), and a confirmatory factor analysis in this data set yielded an excellent fit for a model specifying these three subscales as indicators of a negative emotionality latent variable (Sugimura & Rudolph, 2012). Thus, scores were computed as the mean of the subscales, with higher scores indicating more negative emotionality. Previous studies indicate that parent reports of temperament are reliable (Rothbart, Ahardi, Hershey, & Fisher, 2001; Simonds et al., 2007) and stable (Rothbart, Ahadi, Hershey, & Fisher, 2001), and correlate with child reports (Simonds et al., 2007), behavioral observations (Wilson, 2006), and laboratory tasks (Simonds et al., 2007).

Child victimization

At W₁, children completed a revised version (Rudolph et al., 2011) of the Social Experiences Questionnaire (Crick & Grotpeter, 1996), which assesses overt victimization (being the target of behaviors intended to harm through physical damage or the threat of such damage; e.g., “How often do you get pushed or shoved by another kid?”) and relational victimization (being the target of behaviors intended to harm through the manipulation of peer relationships; e.g., “How often does another kid tell lies about you to make other kids not like you anymore?”). Eleven items were added to the original measure to provide a more comprehensive assessment of victimization, resulting in a 22-item measure. Children rated how often they experienced each type of victimization on a 5-point scale (Never to All the time). Scores were computed as the mean of the items. Children’s self-reports of victimization provide valid information that corresponds with peer (e.g., Graham & Juvonen, 1998) and parent (Bollmer, Harris, & Milich, 2006) reports and behavioral observations (e.g., Kochenderfer & Ladd, 1997). The mean victimization score (Table 1) corresponds to a response of “Almost Never” on the rating scale, as would be expected in a nonselected community sample and similar to reported scores in comparable early elementary school samples using the Social Experiences Questionnaire (e.g., among children of average social status, M = 2.25–2.39, Crick & Grotpeter, 1996; M = 2.16–2.27, Greisbrecht, Leadbeater, & MacDonald, 2011).

Results

Correlational analyses

To assess the degree of overlap among the predictors, we examined zero-order correlations among maternal emotional functioning (emotional awareness and depression), family demographics, and child characteristics (Table 2). Maternal emotional awareness was negatively associated with maternal depression and child negative emotionality and was positively associated with marital status. Maternal depression was negatively associated with marital status and family income and was positively associated with child negative emotionality. Family income was positively associated with marital status and negatively associated with child negative emotionality and child victimization. Significant correlations were small to moderate in size, suggesting that maternal emotional functioning, family demographics, and child characteristics were associated yet distinct.

Table 2.

Intercorrelations of predictors at Wave 1 (N = 330).

Measure	1	2	3	4	5	6	7
1. Maternal emotional awareness	–	−.48***	.11*	.10†	−.29***	−.01	−.03
2. Maternal depression		–	−.11*	−.25***	.26***	−.06	.09
3. Maternal marital status (0 = not married, 1 = married)			–	.58***	−.09	−.07	−.07
4. Family income				–	−.17**	−.10†	−.14*
5. Child negative emotionality					–	−.01	.02
6. Child sex (0 = boys, 1 = girls)						–	.09†
7. Child victimization							–

^†p < .10;

^*p < .05;

^**p < .01;

^***p < .001.

To assess the independent association of family demographics, child characteristics, and maternal emotional functioning with socialization of coping, we examined zero-order correlations concurrently and over time (Table 3). W₁ marital status and family income were positively associated with W₂ primary control engagement and negatively associated with W₂ cognitive avoidance. W₁ child negative emotionality was negatively associated with W₁ primary control engagement and W₂ cognitive restructuring and positively associated with W₁ behavioral avoidance and W₂ cognitive avoidance. Child victimization was negatively associated with W₂ primary control engagement. Child sex was not significantly associated with socialization of coping at either wave.

Table 3.

Correlations between predictors and socialization of coping.

Measure	W1 coping suggestions (N = 330)				W2 coping suggestions (N = 330)
	Primary control engagement	Cognitive restructuring	Cognitive avoidance	Behavioral avoidance	Primary control engagement	Cognitive restructuring	Cognitive avoidance	Behavioral avoidance
W1 marital status (0 = not married, 1 = married)	.07	.01	−.04	.08	.12*	.06	−.15**	−.00
W1 family income	.04	−.02	.00	.04	.13*	−.01	−.11*	.01
W1 child negative emotionality	−.14*	−.06	.10†	.14*	−.03	−.15**	.14*	−.01
Child sex (0 = boys, 1 = girls)	.07	.05	−.06	−.06	.06	.03	−.03	−.07
W1 child victimization	−.03	−.07	.07	.02	−.12*	.07	.03	.04
W1 maternal emotional awareness	.33***	.06	−.31***	−.05	.29***	.16**	−.31***	−.10†
W1 maternal depression	−.17**	−.15**	.17**	.19**	−.18**	−.24***	.26***	.12*

W₁: Wave 1; W₂: Wave 2.

^*p < .05;

^**p < .01;

^***p < .001;

^†p < .10.

As expected, W₁ maternal emotional awareness was positively associated with primary control engagement at both waves and W₂ cognitive restructuring and was negatively associated with cognitive avoidance at both waves. Also as expected, W₁ maternal depression was negatively associated with primary control engagement and cognitive restructuring at both waves and was positively associated with cognitive avoidance and behavioral avoidance at both waves.

In sum, correlation analyses revealed small to moderate associations among family demographics, child characteristics, and maternal emotional functioning as well as significant associations between family demographics, child characteristics, and maternal emotional functioning and socialization of coping. Together, these findings support the importance of examining the unique contribution of maternal emotional functioning to socialization of coping, above and beyond the effects of family demographics and child characteristics.

Hierarchical multiple regression analyses

Hierarchical multiple regression analyses were conducted to examine the prospective contribution of maternal emotional functioning to socialization of coping. W₁ family demographics, child characteristics, and maternal emotional functioning were entered to predict the four dimensions of W₂ socialization of coping, adjusting for earlier socialization of coping (Table 4). For each socialization of coping dimension, W₁ socialization of coping and family demographics were entered in the first step, child characteristics were entered in the second step, maternal emotional functioning was entered in the third step, and the interaction between maternal emotional functioning and child sex was entered in the fourth step. This approach provided a conservative test that enabled us to examine the unique contribution of maternal emotional functioning beyond other predictors.

Table 4.

Hierarchical multiple regression analyses predicting W₂ socialization of coping (N = 327).

W1 predictor variable	W2 primary control engagement			W2 cognitive restructuring			W2 cognitive avoidance			W2 behavioral avoidance
	β	t	ΔR2	β	t	ΔR2	β	t	ΔR2	β	t	ΔR2
Step 1			.31***			.11***			.09***			.21***
W1 socialization of coping	.54	11.58***		.33	6.19***		.26	4.88***		.45	9.12***
Marital status (0 = not married,1 = married)	.02	.36		.09	1.36		−.11	−1.69†		−.04	−.57
Family income	.10	1.74†		−.06	−.86		−.05	−.75		.01	.11
Step 2			.01			.03*			.01			.01
Child negative emotionality	.08	1.58		−.13	−2.55*		.10	1.81†		−.08	−1.60
Child sex (0 = boys, 1 = girls)	.04	.89		.01	.15		−.03	−.49		−.04	−.77
Child victimization	−.08	−1.70†		.09	1.69†		−.02	−.30		.03	.64
Step 3			.02*			.03**			.05***			.01
Maternal emotional awareness	.13	2.33*		.04	.59		−.17	−2.75**		−.10	−1.79†
Maternal depression	−.02	−.35		−.18	−2.97**		.12	2.00*		−.00	−.01
Step 4			.01			.01			.02†			.02*
Maternal emotional awareness × child sex	.08	.94		.16	1.78†		−.22	−2.39*		.04	.43
Maternal depression × child sex	−.09	−1.20		.04	.49		−.11	−1.24		.22	2.65**

W₁: Wave 1; W₂: Wave 2.

Note. β = standardized regression coefficient at each step. ΔR² = change in R² for each step.

^†p < .10;

^*p < .05;

^**p < .01;

^***p < .001.

Significant maternal emotional functioning × sex interactions were further examined in two ways. First, regression analyses were conducted to determine the significance of the simple slopes within sex; results were graphed at 1 SD above and below the mean on emotional functioning. Second, regression analyses were conducted to determine the effect of sex at low and high levels of emotional functioning (i.e., to examine whether the end points of plotted interactions differed; Aiken & West, 1991). Specifically, to evaluate whether the coping suggestions of mothers high in emotional functioning were significantly different for girls and boys, we repeated our central regression analyses but centered maternal emotional functioning at 1 SD above the mean. To evaluate whether the coping suggestions of mothers low in emotional functioning were significantly different for girls and boys, we repeated our central regression analyses but centered maternal emotional functioning at 1 SD below the mean. For descriptive purposes, we note marginally significant effects but do not interpret these effects.

Predicting primary control engagement suggestions

After adjusting for W₁ primary control engagement suggestions, family income marginally predicted more W₂ primary control engagement suggestions. At step two, child victimization marginally predicted fewer W₂ primary control engagement suggestions. Consistent with hypotheses, at step three, maternal emotional awareness made a significant positive incremental contribution. At step four, maternal emotional functioning × child sex interactions were not significant.

Predicting cognitive restructuring suggestions

After adjusting for W₁ cognitive restructuring suggestions, family demographics did not significantly predict W₂ cognitive restructuring suggestions. At step two, child negative emotionality predicted fewer cognitive restructuring suggestions and child victimization marginally predicted more cognitive restructuring suggestions. Consistent with hypotheses, at step three, maternal depression made a significant negative incremental contribution. At step four, the maternal emotional awareness × child sex interaction was marginally significant.

Predicting cognitive avoidance suggestions

After adjusting for W₁ cognitive avoidance suggestions, marital status marginally predicted fewer W₂ cognitive avoidance suggestions. At step two, child negative emotionality marginally predicted more W₂ cognitive avoidance suggestions. Consistent with hypotheses, at step three, maternal emotional awareness made a significant negative incremental contribution and maternal depression made a significant positive incremental contribution. At step four, the maternal emotional awareness × child sex interaction was significant. Decomposition of this interaction revealed that maternal emotional awareness predicted fewer cognitive avoidance suggestions for girls (β = −.28, t = −3.50, p < .01) but not boys (β = −.08, t = −.79, NS; see Figure 1(a)). Further analyses examining whether the end points of the plotted interaction differed at high or low emotional awareness (see earlier description) revealed that mothers high in emotional awareness made fewer cognitive avoidance suggestions to girls than boys (β = −.17, t = −2.10, p < .05), whereas mothers low in emotional awareness did not differentially suggest cognitive avoidance across child sex (β = .12, t = 1.52, NS).

Figure 1.

Effects of (a) W₁ maternal emotional awareness on W₂ cognitive avoidance suggestions and (b) W₁ maternal depression on W₂ behavioral avoidance suggestions, moderated by child sex. Analyses adjust for W₁ socialization of coping. W₁: Wave 1; W₂: Wave 2.

Predicting behavioral avoidance responses

After adjusting for W₁ behavioral avoidance suggestions, family demographics and child characteristics did not significantly predict W₂ behavioral avoidance suggestions. At step three, maternal emotional awareness made a marginally significant negative incremental contribution. At step four, the maternal depression × child sex interaction was significant. Decomposition of this interaction revealed that maternal depression predicted marginally more behavioral avoidance suggestions for girls (β = .15, t = 1.87, p < .10) and fewer behavioral avoidance suggestions for boys (β = −.17, t = −1.99, p < .05; see Figure 1(b)). Further analyses examining whether the end points of the plotted interaction differed at high or low depression (see earlier description) revealed that mothers low in depression made fewer behavioral avoidance suggestions to girls than boys (β = −.19, t = −2.52, p < .05), whereas mothers high in depression did not differentially suggest behavioral avoidance across child sex (β = .11, t = 1.46, NS).

Discussion

Despite the importance of mothers’ coping suggestions (Abaied & Rudolph, 2010b; 2011), the antecedents of coping socialization remain relatively unknown (for one exception, see Abaied & Rudolph, 2010a). This study investigated whether maternal emotional functioning contributes to the coping suggestions mothers make to children in the context of a common childhood stressor. Results indicated that maternal emotional awareness and depression made unique contributions to specific domains of socialization of coping over time and after considering other characteristics of the family and child. These results inform and refine models of the determinants of parenting and have practical implications for parenting interventions.

We hypothesized that mothers with adequate emotional, behavioral, motivational, and cognitive resources (reflected in high emotional awareness or low depression) would encourage their children to engage with stress (either by addressing the source of stress or their emotions or by changing the way they think about a stressful situation), whereas mothers with compromised resources (reflected in low emotional awareness or high depression) would encourage their children to avoid stressors or their responses. In support of our hypotheses, maternal emotional awareness predicted more primary control engagement and fewer cognitive avoidance suggestions, whereas maternal depression predicted fewer cognitive restructuring suggestions and more cognitive avoidance suggestions over time. Furthermore, mothers with adequate resources differed in the avoidance suggestions they made to girls and boys, whereas mothers with compromised resources did not. Specifically, mothers high in emotional awareness made fewer cognitive avoidance suggestions to girls than boys, and mothers low in depression made fewer behavioral avoidance suggestions to girls than boys.

Maternal emotional functioning as a predictor of socialization of coping

Emotional functioning may influence socialization of coping by enhancing or depleting the resources available to mothers to devote toward this task. Mothers with adequate resources (high emotional awareness or low depression) may make more active, engagement coping suggestions because they are better equipped to help their children use such strategies. Gohm and Clore (2002) theorized that in times of stress, individuals high in emotional awareness use fewer resources to process their emotions and thus can allocate additional resources toward engaging in active forms of coping. In contrast, individuals with depression may have more difficulty processing emotions (Rude & McCarthy, 2003; Salovey et al., 1995) and using active coping strategies (Churchill et al., 2010). High levels of emotional awareness and low levels of depression may similarly provide mothers with resources to allocate toward encouraging active forms of coping (i.e., primary control engagement or cognitive restructuring), whereas low levels of emotional awareness or high levels of depression may undermine mothers’ ability to encourage active coping. Instead, mothers with limited resources may find it easier to encourage their children to orient thoughts away from stress (i.e., cognitive avoidance) as these suggestions do not require the mothers to process and engage with their children’s emotions.

Mothers with high emotional awareness and low depression also may show more parenting flexibility. Just as having adequate psychological resources (reflected in low depression) allows mothers to display contingent responsiveness when interacting with their child (Hoffman & Drotar, 1991; Johnston et al., 2002), it may similarly allow mothers to modify their socialization efforts based on their beliefs about what strategies will be most beneficial for their child, given his or her sex. Mothers with sufficient resources may be less likely to suggest cognitive avoidance to girls than to boys due to beliefs that girls are more able and/or inclined to process and express emotions (Fivush et al., 2000). Mothers with adequate resources also may be more likely to suggest behavioral avoidance to boys than girls due to concerns that boys’ active engagement in peer stress could manifest itself in aggression. For example, Abaied and Rudolph (2010b) found that mothers’ engagement coping suggestions predicted externalizing problems in boys, but not in girls, exposed to high levels of peer stress. In contrast, mothers with compromised resources may be less flexible in their socialization of coping efforts, perhaps stemming from diminished awareness of children’s abilities or needs; rather, their suggestions may be driven more by their own depleted resources than by the sex of their child.

Of note, maternal emotional awareness and depression made both shared and unique contributions to socialization of coping. Zero-order correlations revealed that both were independently associated with each domain of socialization of coping at W₂. However, when emotional awareness and depression were considered together in regression analyses, maternal emotional awareness uniquely predicted more primary control engagement, whereas maternal depression uniquely predicted less cognitive restructuring; both predicted cognitive avoidance (in the opposite direction). Because emotional awareness encompasses valuing and attending to emotion, as well as expressing emotion, this aspect of emotional functioning may be especially important in predicting mothers’ primary control engagement suggestions. Primary control engagement coping suggestions in part address children’s emotions and encourage emotional expression, which parallel the abilities and inclinations of individuals with heightened emotional awareness. Maternal depression may be particularly important in predicting cognitive restructuring suggestions. These suggestions require positive thinking (e.g., encouraging their child to look for something good in what happened), a thought pattern that stands in direct contrast to cognitive and emotional sequelae of depression (e.g., pessimism, lack of positive affect). Thus, different aspects of emotional functioning may exert both common and unique influences on socialization of coping.

Contributions to theory and research on determinants of parenting

This study supports theoretical models positing that parent psychological characteristics (Belsky, 1984; Conley et al., 2004) and affective experiences (Dix, 1991) are key determinants of parenting. Prior theory and research on the role of emotion in parenting suggest that specific emotions organize and direct parenting by influencing parents’ cognitive appraisals, style of communication, and child-directed approach or avoidance behaviors (Dix, 1991). The present results suggest that just as specific emotions can orient parents’ attention and actions, mothers’ general emotional functioning also guides parenting behavior.

This study is the first to our knowledge to examine emotional awareness as a predictor of parenting. Bridging and extending prior research indicating that emotional awareness is associated with empathy (Salovey et al., 2002) and active forms of coping (Gohm & Clore, 2002; Salovey et al., 2002), this study revealed that emotional awareness guides mothers’ socialization of coping. Although there is substantial research on depression as a predictor of parenting, this study extends previous research by demonstrating that maternal depression predicts not only general parenting styles, such as the display of positive or negative affect and sensitivity (Goodman & Gotlib, 1999; Lovejoy et al., 2000), but also specific socialization styles implicated in the development of children’s coping and adjustment.

Family and child characteristics as predictors of socialization of coping

In line with theories suggesting that child characteristics elicit parenting behavior (e.g., Bell, 1968; Belsky, 1984), mothers of children high in temperamental negative emotionality reported making fewer cognitive restructuring suggestions. Prior research indicates that children high in negative emotionality use fewer cognitive restructuring strategies (Miller et al., 2009). Thus, mothers may be responding to their child’s inclinations, or they may believe that their child does not possess the necessary skills to engage in these strategies. Mothers with a child who has strong negative reactions to stress and is difficult to soothe also may make fewer cognitive restructuring suggestions because they assume strategies such as positive thinking will not be adequate to help their child cope with stress (Compas, Connor-Smith, & Jaser, 2004). These results suggest that not only mothers’ but also children’s resources influence parenting behavior. Moreover, the influence of temperament on parents’ coping suggestions may be important to consider in future research as there is evidence that children’s coping strategies may mediate the association between temperament and depression (for a review, see Compas, Connor-Smith, & Jaser, 2004).

Family demographics and child victimization did not significantly contribute to socialization of coping over time. As anticipated, zero-order correlations revealed that being married and having a higher family income were associated with more primary control engagement suggestions and fewer cognitive avoidance suggestions at the second wave. However, these associations were reduced to marginal or nonsignificance in regression analyses after adjusting for prior socialization of coping, suggesting that family demographics are important but do not predict changes in mothers’ socialization over time. Child victimization was largely unassociated with socialization of coping in both correlation and regression analyses. The resources available to mothers and their children may make mothers more or less able or inclined to suggest particular coping strategies regardless of the child’s level of peer victimization.

Implications, limitations, and future directions

This study makes a significant novel contribution to theory and research on the individual differences in parenting by illustrating how mothers’ emotional functioning shapes how they guide their children’s coping. Although the contributions of emotional functioning to socialization of coping were small, it is important to note that the analyses used a conservative approach in which these effects were examined over an extended period of time after accounting for prior socialization of coping, family demographics, and child characteristics. Thus, these effects, which were consistent with our hypotheses, were quite robust.

To build further on these contributions, research is needed on the context-specificity of socialization of coping and its antecedents and consequences. This study examined socialization of coping in response to peer victimization, as this is a prominent and influential form of stress in children’s lives. Yet mothers’ responses to their child’s stress, as well as the antecedents and consequences of socialization of coping, may vary depending on the type of stress the child faces. For example, a child may benefit from primary control engagement suggestions when dealing with controllable stressors, such as struggling with academics, but may benefit from disengagement suggestions in uncontrollable, potentially dangerous situations, such as neighborhood violence (Band & Weisz, 1988; Gonzales, Tein, Sandler, & Friedman, 2001). Indeed, research suggests that the consequences of socialization of coping depend on the level and type of stress to which children are exposed (Abaied & Rudolph, 2010b). Mothers with adequate resources may be particularly in tune with the consequences of various forms of coping and able to determine what forms of coping are most effective in a given situation.

As the association between maternal emotional functioning and socialization of coping is further explored, the use of alternate methods of assessment also would be beneficial. This study relied on mothers’ reports of their own emotional functioning and socialization of coping. Maternal reports of socialization of coping are associated with children’s self-reported coping (Abaied & Rudolph, 2011; Kliewer et al., 1996; Shipman & Zeman, 2001) and teacher reports of children’s coping (Eisenberg et al., 1996), thus establishing their validity. However, to ensure that results are not inflated by shared method variance, socialization of coping could be measured through other methods such as observations, child report, or daily diaries. These methods also could provide insight into how mothers use multiple forms of socialization of coping in combination and how socialization of coping varies across contexts. It also will be important to consider the role of fathers in socialization of coping. This study focused solely on mothers’ socialization of coping, but reports of fathers’ responses to their children’s stress could reveal different associations with emotional functioning, unique consequences for children, and interactions with mothers’ responses. Finally, although we included several family characteristics in this study, it will be important to examine whether mothers’ emotional functioning predicts their socialization of coping after accounting for other family characteristics, such as the quality of parent–child interactions (Miller, Kliewer, & Partch, 2010), and to replicate the current findings in a more representative sample to expand the generalizability of the results.

Research is needed to explicitly examine how these results fit with prior work on socialization of coping. Prior research indicates that maternal adult attachment contributes to mothers’ coping suggestions (Abaied & Rudolph, 2010a). Because attachment is broadly related to the regulation of emotion, it is possible that the specific dimensions of emotional functioning examined here underlie the contribution of maternal attachment to socialization of coping. It also would be helpful to explore how adult attachment, mothers’ emotional functioning, and socialization of coping together influence the quality of the parent–child relationship. For example, it may be that when a mother makes certain coping suggestions, such as encouraging her child to express his or her feelings, she conveys to the child that she is emotionally available and values the child’s emotions, fostering a positive mother–child relationship. Thus, maternal attachment and emotional functioning may in part influence the parent–child relationship through mothers’ socialization of coping.

The results of this study can inform interventions designed to optimize parenting. In particular, they suggest that dampened emotional functioning (e.g., depression) guides the coping strategies mothers suggest to their children. Substantial research links maternal depression with child maladjustment (for a review, see Joormann, Eugene, & Gotlib, 2009); a study by Jaser and colleagues (2008) indicates that children’s coping may in part underlie this association. By addressing depressed mothers’ socialization of coping, interventions may promote better coping in children of depressed mothers, deterring future adjustment problems. Looking beyond socialization of coping, this study also has broad implications for parenting interventions. Because mothers’ emotional functioning shapes their engagement in certain parenting behaviors, it would be helpful to assess emotional awareness and depressive symptoms prior to training mothers to implement suggested parenting practices. This study also suggests that although emotional awareness and depression are inversely associated, each exerts distinct influences on parenting behaviors. Parenting interventions in depressed mothers would therefore benefit from considering and addressing the unique influences of emotional awareness and depressive symptoms on parenting.